用户访问网站,Web日志中会记录下大量的用户访问信息,通过挖掘这些日志数据可以获得相关页面、相似用户群体和用户访问模式等信息,Web日志挖掘对于优化网站结构、提供个性化服务和构建智能化网站具有重要作用.数据预处理是保证Web日志挖掘质量的重要基础,预处理主要包括数据清洗、用户识别、会话识别、路径补充和结果评价等工作.概述了数据预处理技术,重点介绍了目前常用的会话划分算法和评价标准,并实现了一个数据预处理系统.
[1] 韩家炜, 孟小峰, 王静, 李盛恩, HAN Jia-Wei, MENG Xiao-Feng, WANG Jing, LI Sheng-En. Web挖掘研究 [J]. 计算机研究与发展 2001.
[2] CHAKRABARTI S. Data mining for hypertext:A tutorial survey [J]. SIGKDD Explorations 2000, 1(02).
[3] PERKOWITZ M, ETZIONI O. Adaptive sites: Automatically learning from user access patterns [A]. California 1997.
[4] PITKOW J. In search of reliable usage data on the WWW [A]. California 1997.
[5] GRAHAM-CUMMING J. Hits and miss-es: A year watching the Web [A]. California 1997.
[6] ZAIANE O R, WASEDA M, HAN J. Discovering Web access patterns and trends by applying OLAP and DATA mining techonlogy on Web logs [A]. Los Alamitos, Santa Babara, CA 1998.
[7] PALIOURAS G, PAPATHEODOROU C, KARKALETSIS V. Clustering the users of large Web sites into communities [A]. San Francisco: Morgan Kaufmann Publishers 2000.
[8] NANOPOULOS A, MANOLOPOULOS Y. Mining patterns from graph traversals [J]. Data and Knowledge Engineering 2001, 37(03).
[9] BAGLIONI M, FERRARA U, ROMEI A. Preprocessing and mining Web log data for Web personalization [A]. Pisa, Italy 2003.
[10] SPILIOPOULOU M, MOBASHER B, BERENDT B. A framework for the evaluation of session reconstruction heuristics in Web-usage analysis [J]. Informs Journal on Computing 2003, 15(05).
